Antiretroviral therapy attenuates disease progression in those infected with human immunodeficiency virus (HIV) but cannot eradicate the infection. Drug sanctuaries are anatomical compartments that harbor the virus while allowing limited drug penetration. Protease inhibitors and non-nucleoside reverse transcriptase inhibitors appear in semen at a fraction (typically, 2 - 15%) of their concentration in blood. In spite of this, we propose that these low concentrations in semen are due to pharmacokinetic phenomena that do not reduce the concentration of active drug. There are two reasons to believe this is the case. First, the structure of the blood-semen barrier is such that it should be permissive to drug passage; only the testes which produces 10% of semen by volume is a true exclusionary barrier. Second, reduced plasma protein binding in semen results in low concentrations in seminal plasma. The unbound, active drug is at effective concentrations in blood and semen. Our first objective is to clarify the role of plasma protein binding as a determinant in the distribution of drugs into the seminal compartment. Blood and seminal plasma samples will be obtained from patients taking any of three protease inhibitors to measure the binding of these antiretrovirals ex vivo and to determine their in vivo distribution into seminal plasma. Seminal viral forms have been shown to be evolutionarily distinct from blood viral forms. Although the testes form only a small portion of the semen by volume, they are the predominant source of seminal lymphocytes. We propose that the testes are the source of the phylogenetically distinct seminal virus forms. Our second objective will be to determine the source of seminal virus. Semen samples will be collected for viral cloning and sequencing from HIV infected patients with and without vasectomies. Those without vasectomies should show a viral form distinct from that of blood plasma consistent with a testicular source. Those with vasectomies should have virus consistent with the blood forms. These results will be important for determining which drugs are effective for suppressing virus in semen and what is the source of seminal virus. This is crucial because the semen is not only a potential drug sanctuary, it is the most common vector for the transmission of infection making it a crucial site for viral suppression.